- Email: [email protected]
Pollen Viability and Pollen-tube Growth Following Controlled Pollination and their Relation to Low Fruit Production in Teak (Tectona grandisLinn. f.)
Annals of Botany 80 : 401–410, 1997
Pollen Viability and Pollen-tube Growth Following Controlled Pollination and their Relation to Low Fruit Production in Teak (Tectona grandis Linn. f.) S U W A N T A N G M I T C H A R O EN* and J O H N N. O W E NS†‡ * ASEAN Forest Tree Seed Centre, Muak Lek, Saraburi, Thailand 18180 and † Centre for Forest Biology, Uniersity of Victoria, Victoria, B.C., Canada V8W 2Y2 Received : 28 August 1996
Accepted : 2 April 1997
Pollen released at 1100 h has the highest viability (92±2 %) but is no longer viable 3 d (84 h) after anthesis. In itro pollen-tube growth is fast (140 µm h−") and increases significantly within the first 8 h. In io pollen tubes also grow quickly and reach the base of the style within 2 h after pollination and enter the micropyle 8 h after pollination. There is no significant difference between self- and cross-pollination in either the rate and the number of pollen tubes in the pistil and the number of ovules penetrated by a pollen tube. Teak has late-acting gametophytic self-incompatibility ; the majority of pollen tubes grow through the style but some do not continue to grow from the style towards the embryo sacs. Pollen-tube abnormalities include swollen, reversed, forked and tapered tips and irregular and spiralling tubes. These are most prevalent in self-pollination (20±4 %). The index of self-incompatibility of 0±17 and low fruit set following self-pollination (2±49 %) indicates that teak is mostly self-incompatible. Drastic fruit abortion occurs within the first week following controlled pollination. Within 14 d, fruit size and fruit set from cross-pollination is generally much greater than from self-pollination. # 1997 Annals of Botany Company Key words : Tectona grandis, pollen viability, pollen-tube growth, pollination, controlled pollinations, incompatibility.
INTRODUCTION Like many hermaphroditic species (Stephenson, 1981 ; Willson and Burley, 1983 ; Sutherland, 1987), teak (Tectona grandis Linn. f.) produces low fruit set due to selfincompatibility (SI) (Hedegart, 1973 ; Tangmitcharoen and Owens, 1997). This is also true for the two other members of the Verbenaceae ; Vitex cooperi and Gmelina arborea (Bolstad and Bawa, 1982 ; Bawa, Perry and Beach, 1985). In teak, it was suggested that insect pollinators foraged mostly on the same tree resulting in a high level of selfing which led to lower fruit set and production of self-fruits with low germination rates (Hedegart, 1973). There have been several studies of the reproductive biology and pollination of teak (Cameron, 1968 ; Bryndum and Hedegart, 1969 ; Hedegart, 1973 ; Egenti, 1974 ; Kedarnath, 1974 ; Tangmitcharoen and Owens, 1997). Bryndum and Hedegart (1969) described the procedure for controlled pollination using small bags to isolate single flowers and large bags to isolate an entire inflorescence. Hand crosspollination was reported to increase the fertilization percentage by 6 to 60 % (average 20 %) compared to 0±4 to 5±1 % (average 1±3 %) from open-pollination. However, there has been no report on the pattern of pollen-tube growth or quantification of pollen tubes within the pistil following hand-pollinations and their role in fruit production of teak. To ensure the success of hand-pollinations, it is important to determine pollen viability prior to pollen application. In general, pollen viability indicates the ability to deliver the ‡ For correspondence. Fax 604 721 6611, e-mail : jowens!uvic.ca
0305-7364}97}10040110 $25.00}0
sperm cells to the embryo sac after pollination (Shivanna, Linskens and Cresti, 1991). In species having gametophytic self-incompatibility (GSI), pollen tends to be binucleate and germinate readily in itro (Brewbaker and Majumder, 1961 ; Brewbaker, 1967). Teak pollen is reported to be binucleate (Siripatanadilox, 1974) but speed or ease of germination are not known. Mulcahy and Mulcahy (1983) also suggested that pollen tubes from binucleate pollen tend to be inhibited in the style in GSI. In angiosperms, the rate of in itro pollen-tube growth varies from 60–20 000 µm h−". This is at least ten times faster than gymnosperm pollen (Hoekstra, 1983). Brewbaker and Majumder (1961) proposed that in itro growth of binucleate pollen tubes is approximately 10 % of in io pollen-tube growth due to limited reserve food in the pollen. However, Bryndum and Hedegart (1969) reported that teak in itro pollen-tube growth rate was high, reaching a maximum of 9±3 mm in 24 h. Self-incompatibility may occur in the stigma, style or ovary (Seavey and Bawa, 1986) but inhibition in the style is thought to be more characteristic of GSI (de Nettancourt, 1977). In open-pollination of teak (Tangmitcharoen and Owens, 1997) pollen-tube inhibition from GSI occurred anywhere along the pistil, but was mostly in the lower portion of the ovary. Incompatible pollen-tube arrest within the ovary, described as late-acting SI, has now been reported to occur primarily in woody species (Seavey and Bawa, 1986 ; Sage, Bertin and Williams, 1994). This study reports on pollen viability and longevity, the patterns and the rates of in itro and in io pollen-tube growth following controlled pollinations, various abnormalities observed in io pollen-tube growth, rate and form
bo960440
# 1997 Annals of Botany Company
402
Tangmitcharoen and Owens—Controlled Pollination and Fruit Production in Teak
of SI, reproductive success, fruit set and rate of fruit abortion of teak.
Three 40-year-old plantation teak trees growing at Muaklek, Saraburi, Thailand, (14° 40« N, 101° 17« E) were used. Two or three trees were used for each part of the study. Scaffolding was erected to a height of 8 m adjacent to each tree. Flowering occurs during the rainy season from June to August and the most abundant flower production occurs in July and August The study began in July 1994.
previously bagged flowers and was applied by tapping the anther, held with forceps, over the stigma of a receptive flower between 1100 and 1200 h. Pollen donors were chosen from trees located nearby (5 m) and further (500 m) away. For self-pollinations, pollen from the same tree was used and applied as described above. Since approx. 1 to 3 % of the flowers in a given inflorescence open each day (Tangmitcharoen and Owens, 1997), hand pollinations were carried out on successive days for each inflorescence. The bags were removed the day after all of the open flowers in the bag were pollinated to prevent inflorescence damage by heavy rains and strong winds against the bags.
Pollen germination and pollen-tube growth
Microscopy
Preliminary studies indicated that media containing only sucrose may not result in optimum pollen germination. Therefore, Brewbaker’s solution (Brewbaker and Kwack, 1963) was used for this study. To determine pollen germination, mature pollen was extracted at 12 different times (0, 2, 4, 6, 8, 10, 12, 24, 36, 48, 72, 84 h) after flower opening (0700 h). For each collection, two flowers per inflorescence (four inflorescences per tree) were collected from each of the three trees. The fresh pollen from each flower was then placed in a petri dish containing 10 % Brewbaker’s solution with 10 % sucrose for 4 h. A drop of this mixture was placed on a microscope slide and covered with a cover slip. Using a light microscope, approx. 200 pollen grains per sample were observed to determine the percentage and the time of pollen germination, i.e. when the pollen tubes were equal or greater in length than pollen diameter. To determine in itro pollentube growth, mature pollen collected at 1300 h was placed on a drop of the same solution and covered with a coverslip. Using an eyepiece gradicule, 90 germinated grains from the three trees were measured for pollen-tube length at 1, 2, 4, 8, 12 and 24 h after sowing.
Specimens used for transmission electron microscopy (TEM) were fixed in 2±5 % glutaraldehyde with 0±075 PO % buffer (pH 7±2) for 2 h at room temperature, rinsed in 0±075 PO buffer, then postfixed in 1 % osmium tetroxide % for 1 h. Specimens were dehydrated in an ethanol series, infiltrated with Spurr’s resin (Spurr, 1969) and cured for 18 h at 60 °C. Semithin sections (1 µm) for light microscopy (LM) were stained with Richardson’s stain (Richardson, Jarrett and Finke, 1960). Ultrathin sections were placed on uncoated 200-mesh copper grids, stained with 2 % aqueous uranyl acetate and 0±2 % lead citrate (Reynolds, 1963) and viewed with a JOEL JEM 1200 EX electron microscope.
MATERIALS AND METHODS Plant material and study site
Pollination experiments In a previous study (Tangmitcharoen and Owens, 1997), pollination success was greater in flowers which opened early and during the peak flowering season than in those which opened at the end of the flowering season. Therefore, hand pollinations were generally performed on those flowers which opened at the beginning or middle of the inflorescence flowering period. At pollination, each flower was tagged with coloured thread and one of the following pollination treatments was carried out : (1) flowers not emasculated, self-pollinated, and bagged ; (2) flowers emasculated, crosspollinated, and bagged ; and (3) flowers not emasculated and unbagged (open-pollination). In the emasculated treatments, the undehisced stamens were removed from the newly opened flowers between 0600 and 0700 h on the day of receptivity. Cellophane bags were then placed over the group of open flowers (approx. ten flowers per inflorescence) to exclude all potential pollinators. Each bag was removed from the receptive flowers for about 15 min for hand pollinations. The pollen was obtained from
Pollen-tube growth and self-incompatibility To determine in io pollen-tube growth, 15 flowers from each of five inflorescences from each of two trees were collected at 1, 2, 4, 6, 8, 12, 24 and 48 h following pollination. The pistils were sliced longitudinally to remove the hairy lateral surfaces of the ovary then fixed in ethanolacetic acid (3 : 1 v}v) for 24 h. After rinsing with water, they were cleared in 8 NaOH at room temperature for 2–4 d, or until most of the tissues became transparent. They were then rinsed in water and stained with 0±1 % decolorized aniline blue in 0±1 K PO modified from Dumas and Knox $ % (1983) and Kenrick and Knox (1985). The number and lengths of pollen tubes in the pistils were examined in approx. 350 flowers using epi-fluorescence microscopy. Fruit set and fruit size To determine fruit set and fruit size following controlled pollination, the variously pollinated flowers were observed and monitored every day for 1 month. The index of selfincompatibility (ISI) was determined by dividing the percentage of fruit set from self-pollination by the percentage of fruit set from cross-pollination (Zapata and Arroyo, 1978). Seed efficiency and reproductie success from openpollination Seed efficiency (SE) was calculated as the number of filled seeds, divided by seed potential, multiplied by 100. To
403
Tangmitcharoen and Owens—Controlled Pollination and Fruit Production in Teak
Statistical analysis Means³s.e. were calculated for all measurements. Arcsine transformations were applied to percentage of pollen germination, the percentage of pollinated flowers, percentage of fruit abortion, and distribution of number of seeds per fruit prior to analysis of variance (ANOVA). The among tree variation in in itro pollen germination, rate of in itro and in io pollen-tube growth, and rate of fruit abortion from controlled pollinations were assessed by ANOVA. The Duncan new multiple range test at P ! 0±05 was used to compare the means for a significant difference among the variables.
1400 d d
d
1200 Pollen tube length (micron)
determine filled seed per fruit, 450 fruits were collected and examined using the cutting tests and x-ray at 25 kv for 60 s. Reproductive success is the product of the fruit to flower ratio (Fr}Fl) and the seed to ovule ratio (S}O) (Weins et al., 1987). Ten inflorescences from each of the three trees were counted to obtain the total number of flowers. From the same inflorescences, the number of flowers that developed into mature fruits were counted.
1000
c
800 600 400
b
200 0
a
1
2
4 8 12 Time after sowing (h)
24
F. 2. In itro pollen-tube growth of teak up to 24 h after sowing. Vertical bars represent s.e. Means of each variable with the same letter are not significantly different at P ! 0±05 as determined by the Duncan new multiple range test.
RESULTS In vitro pollen germination and pollen-tube growth
60
50
40 % of flowers
Pollen released at 1100 h (4 h after anthesis) had the highest viability (92±2 %). Pollen viability gradually decreased after 1100 h and 3 d (84 h) after flower opening, pollen was no longer viable (Fig. 1). There was no significant difference in pollen germination among trees. The peak receptive period for teak flowers was 1100 to 1300 h as reported in an earlier study (Tangmitcharoen and Owens, 1997). Pollen germination began within 1 h of being placed in
30
20
100
a b
b
10
d
34–36
31–33
28–30
25–27
22–24
19–21
16–18
d
d
13–15
60
10–12
0
7–9
c
4–6
c
1–3
% pollen germination
80
Pollen grains per stigma F. 3. Percentage of flowers having different numbers of pollen grains on the stigma of control-pollinated teak flowers ; self (*), n ¯ 49 ; open (+), n ¯ 54 ; cross (8), n ¯ 29.
40 c 20 f f 0
0
2
g
4 6 8 10 12 24 36 48 72 84 Time after anther opening (h)
F. 1. In itro pollen germination of teak pollen at various times after anther opening (n ¯ 144). Vertical bars represent s.e. Means of each variable with the same letter are not significantly different at P ! 0±05 as determined by the Duncan new multiple range test.
Brewbaker’s solution. Pollen-tube growth was fast and increased significantly each hour from 0 to 8 h. Within the first 2 h, the average rate of tube growth was about 160 µm h−". During the 3 to 4 h period it was 280 µm h−". Overall, the rate of tube growth within the first 8 h varied from 60 to 280 µm h−" with an average of 140 µm h−" (Fig. 2). After 8 h, there was no significant increase in pollen-tube growth. The rate of pollen-tube growth varied among trees (P ! 0±001).
404
Tangmitcharoen and Owens—Controlled Pollination and Fruit Production in Teak The average number of pollen tubes in the style varied from 5 to 8. The majority of pollen tubes grew through the style (Fig. 6) but some did not continue to grow from the style towards the embryo sacs (Fig. 7). Pollen-tube growth was fast during the first 4 h. Four h after pollination, pollen tubes of 24 % of self-, 29 % of openand 8 % of cross-pollinated flowers had reached an ovule and about 7 % of pollen tubes from open-pollinated plants had reached a micropyle (Fig. 5 A). The first pollen tubes were observed to reach the embryo sacs 8 h after pollination (Fig. 5 B). Pollen-tube growth was slow between the 8 to 12 h and the 20 to 24 h observation periods. At the 20 to 24 h observation period, 70 to 80 % of pollen tubes from the three types of pollination had reached the ovules, 40 to 60 % of pollen tubes had reached the micropyle, but very few (! 5 %) of pollen tubes from self- and open-pollinated flowers and about 20 % from cross-pollinated plants had reached the embryo sacs (Fig. 5 C). The number of pollen tubes at different positions within the pistil at 20 to 24 h following controlled pollination is shown in Table 1. Of the 84 flowers observed 8 to 24 h after pollination, the number of flowers in which pollen tubes entered the micropyle did not differ significantly among the three types of pollination. The average number of pollen tubes in the various portions of the pistil 20 to 24 h after pollination is shown in Table 1. The number of pollen tubes entering the micropyle per flower varied from 0 to 3 but was not significantly different among the three types of pollination (Table 1). In general, only one micropyle per flower was entered by a pollen tube even though four micropyles exist per flower. Occasionally two, very rarely three and never four micropyles per flower were observed to contain pollen tubes. Observation of the pistil at 28, 32, 36 and 48 h after pollination indicated that there was no increase in the number of pollen tubes entering the micropyles after 24 h.
Stigmatic surface
Pollen Pollen tubes
Style
Ovary Ovule Embryo sac
F. 4. Diagram of teak flower showing the pathway of pollen tubes through the stigma and hollow stylar transmitting tissue (shaded area) to the embryo sacs.
In vivo pollen-tube growth following controlled pollination Of the 225 flowers observed following control self-, openand cross-pollinations, 30 to 50 % (average 35 %³4±16) had pollen adhering to the stigma. The number of pollen grains on the stigma varied from 1 to 38 with an average of 6±2³0±63 (n ¯ 130), but most commonly there were only 1 to 3 grains (Fig. 3). The percentage of pollen adhering to the stigma did not differ significantly between the three types of pollination. The pathway of pollen-tube growth to the embryo sacs is shown in Fig. 4. In all three types of pollination, a few pollen grains did not germinate on the stigma (Table 2). Pollen germinated within 1 h of being placed on a receptive stigma. Generally, self-, open-(natural), and cross-pollen tubes grew at the same rate and there were similar numbers of pollen tubes growing through the pistil for the first 24 h.
Abnormalities of pollen-tube growth In the 132 pistils observed during the 24 h following controlled pollination, most pollen tubes were straight with thin pollen-tube walls and tapered tips. Pollen-tube inhibition on the stigma occasionally occurred in all types of pollination (Fig. 8). Several abnormalities occurred in pollen-tube growth at various positions within the pistil (Table 2). Abnormalities included reversing tubes (Fig. 9),
T 1. Comparison of mean numbers (³s.e.) of pollen tubes entering the stigma, at 25, 50, 75 and 100 % of the style length, and in the oule, micropyle and embryo sac at 20–24 h following pollinations Types of pollination
ANOVA
Distance through pistil
Self-
Open-
Cross-
F values
P
Average
25 % of style 50 % of style 75 % of style 100 % of style Ovule Micropyle Embryo sac
7±8³2±0 7±7³2±0 7±3³2±0 7±3³2±0 6±6³1±9 3±9³1±3 0±2³0±09
6±6³2±5 6±4³2±5 6±4³2±5 6±2³2±5 6±0³2±4 4±3³2±0 0±5³0±22
5±2³1±1 4±9³1±1 4±9³1±1 4±9³1±1 4±6³1±2 2±0³0±63 0±5³0±14
0±45 0±52 0±50 0±39 0±28 0±70 1±28
0±64 NS 0±60 NS 0±62 NS 0±68 NS 0±76 NS 0±50 NS 0±29 NS
6±7 6±6 6±4 6±3 5±8 3±5 0±3
NS, non significant.
Tangmitcharoen and Owens—Controlled Pollination and Fruit Production in Teak 100 2–4 h
A
80 60 40 20 0 25%
50%
75%
100%
ov
mc
em
405
by the percentage fruit set from cross-pollination (14±54) was 0±17, indicating that teak is mostly self-incompatible (Zapata and Arroyo, 1978). Fruit set for self- and openpollinated flowers measured 30 d after pollination differed significantly (P ! 0±01) from cross-pollinated flowers. Only 2±49 % of self-pollinated flowers and 6±54 % of openpollinated flowers set fruit, whereas 14±54 % of crosspollinated flowers set fruit (Table 3). There was no significant difference in percent fruit set among the two trees used for this portion of the study for all types of pollination. The rate of fruit abortion was high during the first 10 d. There was a tendency for a lower percentage of fruit abortion following cross-pollination during the 5 to 20 d period after pollination (Fig. 13).
100 Percent of pollen tubes
8–12 h
B
80
Seed efficiency and reproductie success with openpollination
60
In open-pollination, the seed to ovule (S}O) ratio was 0±33 and the fruit to flower (Fr}Fl) ratio was 0±035. Reproductive success, determined by Fr}Fl multiplied by S}O, was low (0±011). The results of cutting tests and x-ray revealed that most commonly (46±2 %) only one of four ovules per flower developed into a mature seed, 30±9 % contained two seeds per fruit, 8 % contained 3 to 4 seeds and 14±9 % of fruits contained no filled seeds (Table 4). There was no significant difference in seeds per fruit among the three trees.
40 20 0 25%
50%
75%
100%
ov
mc
em
100 20–24 h
C
80
Fruit size during early deelopment stages Fruit size during early development, up to 10 d after pollination, differed significantly (P ! 0±0001) among the types of pollination but not among the two trees included in this part of the study. Self-pollination resulted in smaller fruits. Unfortunately, the high abortion rate due to selfing made fruit size data unavailable after 10 d. Mean fruit size resulting from open- and cross-pollination was essentially the same through the first 28 d of development (Table 5).
60 40 20 0 25%
50%
75% 100% ov mc Distance through style
em
F. 5. Percentage of self- (*), open- (natural) (+), and cross- (8) pollen tubes at different positions in the teak pistil after pollination. A, At 2–4 h, self- (n ¯ 14) ; open- (n ¯ 14) ; cross- (n ¯ 4). B, At 8–12 h, self- (n ¯ 14) ; open- (n ¯ 31) ; cross- (n ¯ 11). C, At 20–24 h, self(n ¯ 17) ; open- (n ¯ 45) ; cross- (n ¯ 9). 25, 50, 75 and 100 % refer to percent of the style length ; ov, ovule ; mc, micropyle ; em, embryo sac.
irregular or spiralling tubes (Fig. 10), and an increase of callose deposits resulting in swelling of the tube tip (Figs 11 and 12). Abnormalities were most prevalent in the selfpollinated flowers (20±4 %) (Table 2). Swelling of the tube tip was most common, accounting for 41±7 % of total abnormal tubes. Fruit set and rate of fruit abortion following controlled pollination The index of self incompatibility (ISI), determined by dividing the percentage fruit set from self-pollination (2±49)
DISCUSSION In vitro pollen germination Binucleate pollen generally germinates readily in culture (Mulcahy and Mulcahy, 1983). This was also true for binucleate teak pollen in which germination began within 1 h of pollen being placed in the medium. Using a 14 % sucrose medium for in itro pollen germination of teak, Egenti (1974) and Bryndum and Hedegart (1969) reported that pollen collected during the 1200–1400 h period on the day of anthesis had the highest percentage of viability compared to that collected from 0730–1000 h and at 1800 h. They also found that 10 % of the pollen was still viable 3 d after anthesis. Using Brewbaker’s solution with 10 % sucrose, results from our studies show some differences to those of previous studies. Germination was best for pollen collected 1 h earlier (1100 h) than in previous studies, but the pollen also remained viable for 3 d.
406
Tangmitcharoen and Owens—Controlled Pollination and Fruit Production in Teak
F 6 and 7. Epi-fluorescence micrographs of pollinated teak pistils within 24 h of pollination. Pistils were stained with decolorized aniline blue to localize callose. Fig. 6. Within 4 h, a large number of pollen germinate on the stigma papillae and produce pollen tubes that grow into the stylar canal (arrowheads). Fig. 7. Arrested pollen tubes in the ovary 24 h after pollination. A dense bundle of pollen tubes (between arrowheads) was generally seen in the lower stylar transmitting tissue but few grew towards the lower portion of the ovary and entered the micropyles. Bars ¯ 0±5 and 0±2 mm in Figs 6 and 7, respectively. P, Pollen ; OV, ovule. F 8–12. Abnormalities of pollen tubes found at various sites and times following pollination. Fig. 8. Inhibition of a pollen tube on the stigma 12 h after self-pollination. Fig. 9. Reversal of pollen-tube growth (arrowheads) in the style 4 h after self-pollination. Fig. 10. Irregular pollen tube with callose deposition in pollen tube wall (arrowhead) in the upper portion of the ovary 8 h after open-pollination. Fig. 11. Light micrograph showing swelling of pollen tube (arrowhead) in the lower portion of the ovary about 24 h after self-pollination. Fig. 12. Transmission electron micrograph showing the thickened cell wall and small amount of cytoplasm (*) of the tube tip shown in Fig. 11. Bars ¯ 50 mm in Fig. 8 ; 0±3 mm in Figs 9 and 10 ; 20 µm in Fig. 11 ; and 10 µm in Fig. 12. P, Pollen ; ST, stigma ; PT, pollen tube ; OV, ovule ; OW, ovary wall.
In general, most pollen collected on the day of anthesis developed normal pollen tubes and exhibited more uniform pollen-tube growth than pollen collected on or after the
second day. The highest in itro pollen germination (92±2 %) occurred in pollen released at 1100 h. This indicates that the pollen used for control pollination was highly viable.
407
Tangmitcharoen and Owens—Controlled Pollination and Fruit Production in Teak T 2. Number of flowers that had different abnormalities of pollen-tube growth on the stigma and at arious positions within the pistil within 24 h following controlled pollination
100
80
Types of pollination
Stigma* 50 % of style 75 % of style 100 % of style Upper ovary Lower ovary Total number of flowers having abnormal pollen tubes Total number of flowers observed
Self-
Open-
Cross-
4 3 2 0 1R 1R 2S 1R 0 1R 0 0 1R 0 0 1I 3S 1I 0 10 6 5 (20±41 %) (11±11 %) (17±24 %) 49 54 29
* Stigma with pollen adhering but without pollen tube growth. R, Reversing tubes ; S, swollen tube tip ; I, irregular tube.
T 3. Number and mean (³s.e.) percentage of fruit set 30 d after controlled pollination
% Fruit abortion
Distance through pistil
60
40
20
0
5
10 15 20 Days after pollination
25
30
F. 13. Effect of self- (+), open- (D), an cross- (^) pollination on fruit abortion for the 30 d after pollination based on 14 inflorescences on each of the two teak trees.
Types of pollination Tree number Number of inflorescences treated Number of flowers treated Number of fruits set % fruit set Total (% fruit set)
1 2 1 2 1 2 1 2
Self-
Open-
Cross-
3 4 6 4 5 6 30 126 67 52 109 75 1 2 10 1 10 12 3±33 1±59 14±82 1±92 9±17 15±54 2±49a 6±54a 14±54b ³1±61 ³4±65 ³2±77
Means of each variable followed by the same superscript are not significantly different (P ! 0±05) as determined by the Duncan new multiple range test.
Therefore, most of the failure of pollen-tube growth which occurred after control pollination was not due to the lack of pollen vigour and viability, but probably due to incompatibility. In vitro and in vivo pollen-tube growth in controlled pollination There are extreme variations in in io and in itro pollentube growth rate in angiosperms (Hoekstra, 1983). However, there are few reports of in io or in itro pollen-tube growth in tropical forest trees. Rapid in io pollen-tube growth has been reported in Acacia retinotes in which the pollen tubes reached the ovule within 11 h (Kenrick and Knox, 1985). In other cases growth is slow ; it took 10–20 d for pollen tubes of Eucalyptus woodwardii (Sedgley and Smith, 1989) to reach the ovules. In our study, the average rate of in itro pollen-tube growth varied from 60–280 µm h−" which is similar to the majority (19 of the 26) of angiosperm families
T 4. Frequency distribution of percentage fruit haing 0–4 seeds per fruit 30 d after pollination Numbers of seeds per fruit 0
1
2
14±9a 46±2b 30±9c 67 208 139 3±5 3±7 4±3
Percentage of fruits Numbers of fruits s.e.
3 6±22d 28 2±1
4
Total
1±78d 100 8 450 0±69
Means of each variable followed by the same superscript are not significantly different (P ! 0±05) as determined by Duncan new multiple range test.
T 5. Comparison of means (³s.e.) of fruit sizes in mm during early deelopment Types of pollination Days after pollinations 4 7 10 14 21 28
Self-
Open-
Cross-
2±25³0±07 (20)a 2±36³0±06 (20)b 2±45³0±08 (19)b 2±42³0±09 (20)a 3±86³0±12 (20)b 4±2³0±58 (17)b 3±07³0±09 (14)a 5±71³0±13 (19)b 6±41³0±16 (14)b N}A 7±03³0±21 (18) 7±46³0±73 (16) N}A 12±09³0±49 (15) 12±94³0±58 (14) N}A 14±15³0±33 (11) 14±17³0±34 (13)
Numbers in brackets represent the number of fruits observed. Means of each variable followed by the same superscript are not significantly different (P ! 0±05) as determined by the Duncan new multiple range test. N}A, Data not available due to high abortion rate followed selfpollination.
reported by Hoekstra (1983). In our study pollen tubes reached about 1±2 mm in 24 h. This contrasts with studies by Bryndum and Hedegart (1969) for teak where in itro
408
Tangmitcharoen and Owens—Controlled Pollination and Fruit Production in Teak
pollen-tube growth was very fast, up to 9±3 mm with an average of 5±53 mm in 24 h. Although the rate of in itro pollen-tube growth in teak would be considered fast, it was not when compared to the length of pistil. Pistil length averaged 6±55 mm and consequently 24 h after sowing pollen tubes in itro were only about 18 % of the way down the pistil. The in itro growth rates underestimate the in io rates in teak in which some pollen tubes reached the ovules 4 h after pollination. Our results agree with the study by Brewbaker and Majumder (1961) that in itro growth of binucleate pollen tubes was approx. 10 % of in io pollentube growth. Rosen and Gawlick (1966) and Rosen (1971) proposed that in io binucleate pollen-tube growth consists of two phases : autotrophic and heterotrophic. In the first autotrophic phase pollen tubes grow from their own reserves. Pollen grains usually contain enough reserve food to support germination (Johri, 1992). Pollen-tube growth during the first phase is relatively slow and free of callose plugs (Mulcahy and Mulcahy, 1983). Brewbaker (1967) suggested that this phase ends with gamete formation. Pollen-tube growth in itro, or in an incompatible style, may terminate at this point. The second, heterotrophic, phase is observed only in io within a compatible style. During this phase, pollen tubes grow rapidly and form callose plugs, indicating a shift to heterotrophic nutrition (uptake of substances from the pistil). This may account for the limited teak pollentube length in itro using Brewbaker’s media. It may not contain all of the nutrients needed for the transition of pollen tubes to the second phase (Read, Bacic and Clarke, 1992). The continual rapid pollen-tube growth in the stylar canal of teak may be related to movement of the specific proteins from the transmitting tissue, as suggested by Mascarenhas (1993). Hohri (1992) also suggested that a pollen tube growing in a hollow style, as in teak, derives nourishment from the glandular epidermis of the stylar canal and the adjacent tissue becomes depleted of metabolites. Despite these complications, in itro pollen-tube growth measurements reflect some useful pollen properties (Hoekstra, 1983). Based on our observations of in io pollen-tube growth, it is likely that the period for the first phase of in io pollentube growth in teak was very short (probably less than 30 min). Within 1 h following pollination, callose plugs were found near the stigma which corresponds to first phase of growth. The density of pollen deposited on the stigma is reported to affect both pollen germination and rate of pollen-tube growth in some tropical forest trees, including Leucaena leucocephala (Ganeshaiah, Shaankar and Shivashankar, 1986). Our study showed that this may not be true for teak. The density of pollen deposited on the teak stigma did not affect the number of pollen tubes entering the micropyle in any pollination trials. Even when the teak stigma was saturated with pollen (up to 38 grains per stigma), the ovule often remained unpenetrated. There was little pollen competition in the teak style, i.e. an increase in pollen load on the stigma resulted in a proportional increase in the number of pollen tubes penetrating to the base of the ovary.
Pollen-tube inhibition and incompatibility in teak There was no difference in the number or sites of pollentube inhibition among open-, self-, and cross-pollinations in teak until pollen tubes reached the embryo sac (Fig. 5). Also, in the previous report on open-(natural) pollination for teak (Tangmitcharoen and Owens, 1997), the site of the final pollen-tube arrest was generally at the ovary, and most pollen tubes did not enter the embryo sac. Other studies in Eucalyptus woodwardii (Sedgley and Smith, 1989) and Medicago satia (Cooper and Brink, 1940 ; Sayers and Murphy, 1966) have shown that the number of penetrated ovules following self-pollination was less than that following cross-pollination. In angiosperms, self-incompatibility (SI) systems that operate in the ovary, described as late-acting SI systems, were initially assumed to be uncommon in angiosperms (de Nettancourt, 1977). Seavey and Bawa (1986) reported that this notion was erroneous : the scarcity of late-acting SI may have been exaggerated by the preference for studies of breeding systems for small, short-lived, herbaceous plants. Late-acting SI has been reported to be important in the breeding systems of many angiosperms and may be more important in woody angiosperms (Seavey and Bawa, 1986 ; Sage et al., 1994). They were classified into four categories : (1) ovarian inhibition of incompatible pollen tubes before the ovule is reached ; (2) pre-fertilization inhibition in the ovule ; (3) post-zygotic rejection of the embryo ; and (4) ovular inhibition. Late-acting pollen-tube inhibition has been reported in some hardwood species ; Acacia retinodes (Kenrick, Kaul and Williams, 1986), Eucalyptus morrisbyi (Potts and Savva, 1989), and E. woodwardii (Sedgley and Smith, 1989). In these, self- and cross-pollen tubes appear similar through the stigma and style, but self-pollen tubes were inhibited in the ovary. In teak, the late-acting system appears to be associated with inhibition after pollen tubes enter the ovule, however, a post-zygotic stage has not been ruled out. Post-zygotic outcrossing mechanisms have been reported in some tropical woody species including four species of Eucalyptus (Griffin, Moran and Fripp, 1987 ; Ellis and Sedgley, 1992) and mango (Shama and Singh, 1970). In all pollination trials, a variety of pollen-tube abnormalities were found at various locations in the pistil and at different times after pollination. Swollen tips, reversing tips, forked tips, tapered tips, irregular tubes and spiralling tubes were observed. There was a tendency towards higher pollen-tube abnormalities in self-pollinated flowers than open- and cross-pollinated flowers (Table 2). Fruit set following controlled pollinations Low fruit set is typical of many hermaphroditic plant species (Stephenson, 1981 ; Willson and Burley, 1983 ; Sutherland and Delph, 1984 ; Sutherland, 1986, 1987 ; Guitian, 1993). Zapata and Arroyo (1978) classified the ISI as : " 1 ¯ self-compatible ; " 0±2 ! 1 ¯ partially selfincompatible ; ! 0±2 ¯ mostly self-incompatible ; and 0 ¯ completely self-incompatible. The index of ISI was 0±17 in teak, placing it in the category of mostly self-incompatible. This helps explain the high rate of fruit abortion that occurs
Tangmitcharoen and Owens—Controlled Pollination and Fruit Production in Teak after self-pollination, but some self-pollinated flowers can still develop into fruits. Dafni (1992) ranked fruit set percent following selfpollination as : 0–3 % (class 0) ¯ self-incompatible ; 3–30 % (class 1) ¯ slightly self-compatible ; and, " 30 % (class 2) ¯ highly self-compatible. Our study shows the rate of fruit set following self-pollination in teak was 2±49 %, which again classifies teak as a self-incompatible species. Fruit set did not differ significantly between self- and open-pollination which suggests a high incidence of self-pollination in nature or very closely related neighbouring trees. Results from our study support an earlier study by Bryndum and Hedegart (1969) in which less than 1 % of self-pollinated flowers developed into fruits. Evidence from the present and our earlier study (Tangmitcharoen and Owens, 1997) show that teak has late-acting GSI, where the pollen-pistil interaction is genetically controlled by the haploid (gametophytic) genome of each pollen tube as it penetrates the diploid pistil (Sedgley and Griffin, 1989). Low fruit set has also been reported in two other species from the Verbenaceae, Vitex cooperi and Gmilina aborlia (Bolstad and Bawa, 1982 ; Bawa, Perry and Beach, 1985), but in these it was not necessarily attributed to self-incompatibility. In teak, drastic fruit abortion occurred within the first week following control pollination and within 14 d fruit size and fruit set from cross-pollination was generally much greater than from self-pollination. Ackerman (1961) reported that fruit set from self-pollination of Ziziphus (chinese jujubes) was less and fruit had a greater tendency to drop prematurely than from cross-pollination. In general, an increase in pollen load by supplementary pollination leads to an increase in fruit set (Sutherland and Delph, 1984 ; Sutherland, 1987). This was shown in macadamia where an increased number of pollen tubes following crosspollination increased initial fruit set (Ito, Eyre and Cabral, 1983 ; Sedgley et al., 1990). Insufficient insect pollinators and their effectiveness appear to be major causes for limited fruit set in teak (Hedegart, 1973 ; Tangmitcharoen and Owens, 1997). Insects forage mostly among the inflorescences of the same tree or nearby relatives, which contribute to inbred fruits (Hedegart, 1973). This results in a lack of heterozygosity, low fruit set and low germination rate in teak. However, Kertadikara and Prat (1995) suggested that there is high heterozygosity in teak populations and it may be maintained by early elimination of self-material through embryo abortion, and low germination rate. Low fruit set after cross-pollination in many tropical forest species probably results from three factors (Bawa et al., 1985) : an artifact of hand-pollination (Bawa et al., 1985) ; a predetermined abortion rate (Bawa and Webb, 1984) ; and inbreeding depression from close relative crosspollination (Haber and Frankie, 1982). In our study these factors may have contributed to the relatively high (85±46 %) fruit abortion of teak with cross-pollination. Even in crosspollinated flowers, only 30³6±1 % (n ¯ 95) of the flowers had pollen on the stigma and fruit abortion occurred more often if hand-pollination was performed on the flowers which developed at the end of the flowering period in an inflorescence, in particular the outermost flowers of the
409
inflorescence. Also the trees used in our study may be close relatives since they were located in the same plantation. A C K N O W L E D G E M E N TS This study was a part of M.Sc. research of Tangmitcharoen Suwan, funded by CIDA (Canadian Development Agency) with the Canadian Forest Service (Petawawa National Forestry Institute) as Canadian Executing Agency to ASEAN Canada Tree Seed Centre, Muak Lek, Saraburi, Thailand. Research support was also provided by the National Sciences and Engineering Council of Canada, Grant 1982 to JNO. We thank the ASEAN Forest Tree Seed Centre, Saraburi, Thailand for providing field and laboratory facilities. We also thank the Thai Danish Dairy Farm, Saraburi, Thailand for providing the collection site. LITERATURE CITED Ackerman WI. 1961. Flowering, pollination, self-sterility and seed development of chinese jujubes. Proceedings of the American Society for Horticultural Science 77 : 265–269. Bawa KS, Perry DR, Beach JH. 1985. Reproductive biology of tropical lowland rain forest trees. I. Sexual systems and incompatibility mechanisms. American Journal of Botany 72 : 331–345. Bawa KS, Webb CJ. 1984. Flower, fruit and seed abortion in tropical forest trees : implications for the evolution of paternal and maternal reproductive patterns. American Journal of Botany 71 : 736–751. Bolstad PV, Bawa KS. 1982. Self-incompatibility in Gmelina arborea L. (Verbenaceae). Silae Genetica 31 : 19–21. Brewbaker JL. 1967. The distribution and phylogenetic significance of binucleate and trinucleate pollen grains in the angiosperms. American Journal of Botany 54 : 1069–1083. Brewbaker JL, Kwack BH. 1963. The essential role of calcium ion in pollen germination and pollen tube growth. American Journal of Botany 50 : 859–865. Brewbaker JL, Majumder SK. 1961. Incompatibility and the pollen grain. Recent Adances in Botany 2 : 1503–1508. Bryndum K, Hedegart T. 1969. Pollination of teak (Tectona grandis L.f.). Silae Genetica 18 : 77–80. Cameron AL. 1968. Forest tree improvement in New Guinea, 1. Teak. 9th Commonwealth Forest Conference. New Delhi. Cooper DC, Brink RA. 1940. Partial self-incompatibility and collapse of fertile ovule as factors affecting seed formation in alfalfa. Journal of Agricultural Research 60 : 453–472. Dafni A. 1992. Pollination ecology. A practical approach. New York : Oxford University Press. de Nettancourt D. 1977. Incompatibility in angiosperms. Berlin, Heidelberg, New York : Springer-Verlag. Dumas C, Knox RB. 1983. Callose and determination of pistil viability and incompatibility. Theoretical and Applied Genetics 67 : 1–10. Egenti LC. 1974. Preliminary studies on pollinators of teak (Tectona grandis L.f.). Research Paper Forest Series, Federal Department of Forest Research, Nigeria. Ellis MF, Sedgley M. 1992. Floral morphology and breeding system of three species of Eucalyptus, Section Bisectaria (Myrtaceae). Australian Journal of Botany 40 : 249–262. Ganeshaiah KN, Shaankar RU, Shivashankar G. 1986. Stigmatic inhibition of pollen grain germination—its implication for frequency distribution of seed number in pods of Leucaena leucocephala (Lam) de Wit. Oecologia 70 : 568–572. Griffin AR, Moran GF, Fripp YJ. 1987. Preferential outcrossing in Eucalyptus regnans F. Muell. Australian Journal of Botany 35 : 465–475. Guitian J. 1993. Why Prunus Mahaleb (Rosaceae) produces more flowers than fruits. American Journal of Botany 80 : 1305–1309. Haber WA, Frankie GW. 1982. Pollination of Leuhea (Tiliaceae) in Costa Rican deciduous forest : significance of adapted and nonadapted pollinators. Ecology 63 : 1740–1750.
410
Tangmitcharoen and Owens—Controlled Pollination and Fruit Production in Teak
Hedegart T 1973. Pollination of teak (Tectona grandis L.). Silae Genetica 22 : 124–128. Hoekstra FA. 1983. Physiological evolution in angiosperm pollen : possible role of pollen vigour. In : Mulcahy DL, Ottaviano, eds. Pollen : Biology and implications for plant breeding. New York : Elsevier Science. Ito PJ, Eyre JM, Cabral S. 1983. Preliminary study on pollination of five macadamia cultivars. In : Stephenson RA, Gallagher EC, eds. Proceedings of the First Australian Macadamia Research Workshop. Brisbane : Australian Macadamia Society Limited. Johri BM. 1992. Haustorial role of pollen tubes. Annals of Botany 70 : 471–475. Kedarnath S. 1974. Genetic improvement of forest tree species in India. India Journal of Genetics Proceedings of 2nd General Congress Sarrao, New Delhi 34A. Kenrick J, Kaul V, Williams EG. 1986. Self-incompatibility in Acacia retinodes : Site of pollen-tube arrest in the nucellus. Planta 169 : 245–250. Kenrick J, Knox RB. 1985. Self-incompatibility in nitrogen-fixing tree, Acacia retinodes : quantitative cytology of pollen tube growth. Theoretical and Applied Genetics 69 : 481–488. Kertadikara AWS, Prat D. 1995. Genetic structure and mating system in teak (Tectona grandis L.f.) provenances. Silae Genetica 44 : 2–3. Mascarenhas JB. 1993. Molecular mechanisms of pollen tube growth and differentiation. Plant Cell 5 : 1303–1314. Mulcahy GB, Mulcahy DL. 1983. A comparison of pollen tube growth in bi- and trinucleate pollen. In : Mulcahy DL, Ottaviano E, eds. Pollen : Biology and implications for plant breeding. New York : Elsevier Science. Potts BM, Savva M. 1989. Self-incompatibility in Eucalyptus. In : Knox RB, Singh MB, Troiani, eds. Pollination ’88. University of Melbourne, Parkville : Biology Research Centre, pp. 165–170. Read M, Bacic A, Clarke AE. 1992. Pollen tube growth in culture. I. Control of morphology and generative nucleus division in cultured pollen tubes of Nicotiana. In : Ottaviano E et al., eds. Angiosperm pollen and oules. New York : Springer-Verlag, 162–167. Reynolds ES. 1963. The use of lead citrate at high pH as an electron opaque stain in electron microscopy. Journal of Cell Biology 17 : 208–212. Richardson KC, Jarrett L, Finke EH. 1960. Embedding in epoxy resins for ultrathin sectioning in electron microscopy. Stain Technology 35 : 313–323. Rosen WG. 1971. Pistil-pollen interactions in Lilium. In : HeslopHarrison J, ed. Pollen deelopment and physiology. London : Butterworths, 239–254. Rosen WG, Gawlick SR. 1966. Relation of lily pollen tube fine structure to pistil compatibility and mode of nutrition. Sixth International Congress for Electron Microscopy, Kyoto. Maruzen, Tokyo. Electron Microscopy 2 : 313–314.
Sage TL, Bertin RI, Williams EG. 1994. Ovarian and other late-acting self-incompatibility system. In : Williams et al., eds. Genetic control of self-incompatibility and reproductie deelopment in flowering plants. The Netherlands : Kluwer Academic Publishers, 116–140. Sayers ER, Murphy RP. 1966. Seed set in alfalfa as related to pollen tube growth, fertilization frequency, and post-fertilization ovule abortion. Crop Science 6 : 365–368. Seavey SR, Bawa KS. 1986. Late acting self-incompatibility in angiosperms. Botanical Reiew 52 : 195–219. Sedgley M, Bell FDH, Bell D, Winks CW, Pattison SJ, Hancock TW. 1990. Self- and cross-compatibility of macadamia cultivars. Journal of Horticultural Science 65 : 205–213. Sedgley M, Griffin AR. 1989. Sexual reproduction of tree crops. London : Academic Press, 378. Sedgley M, Smith RM. 1989. Pistil receptivity and pollen tube growth in relation to the breeding system of Eucalyptus woodwardii (Symphyomyrtus : Myrtaceae). Annals of Botany 64 : 21–31. Sharma DK, Singh RN. 1970. Self-incompatibility in mango (Mangifera indica L.). Horticultural Research 10 : 108–118. Shivanna KR, Linskens HF, Cresti M. 1991. Pollen viability and pollen vigor. Theoretical and Applied Genetics 81 : 38–42. Siripatanadilox S. 1974. Developmenmt of teak flower (Tectona grandis Linn.). Forest Research Bulletin 31 : Faculty of Forestry, Kasetsart University, Thailand. Spurr AR. 1969. A low viscosity epoxy resin embedding medium for electron microscopy. Journal of Ultrastructure Research 71 : 173–184. Stephenson AG. 1981. Flower and fruit abortion : proximate causes and ultimate functions. Annual Reiew of Ecology and Systematics 12 : 253–280. Sutherland S. 1986. Florax sex ratios, fruit set, and resource allocation in plants. Ecology 67 : 991–1001. Sutherland S. 1987. Why hermaphroditic plants produce many more flowers than fruits : experimental tests with Agae mckeleyana. Eolution 41 : 750–759. Sutherland S, Delph LF. 1984. On the importance of male fitness in plants : patterns of fruit-set. Ecology 65 : 1093–1104. Tangmitcharoen S, Owens JN. 1997. Floral biology, pollination, pistil receptivity, and pollen tube growth of teak (Tectona grandis L.f.). Annals of Botany 79 : 227–241. Wiens D, Calvin CL, Wilson CA, Davern CI, Frank D, Seavey SR. 1987. Reproductive success, spontaneous embryo abortion, and genetic load in flowering plants. Oecologia 71 : 501–509. Willson M, Burley N. 1983. Mate choice in plants : Tactics, mechanisms, and consequences. Princeton, New Jersey : Princeton University Press. Zapata TR, Arroyo MTK. 1978. Plant reproductive ecology of a secondary deciduous tropical forest in Venezuela. Biotropica 40 : 221–230.
Pollen Viability and Pollen-tube Growth Following Controlled Pollination and their Relation to Low Fruit Production in Teak (Tectona grandis Linn. f.) S U W A N T A N G M I T C H A R O EN* and J O H N N. O W E NS†‡ * ASEAN Forest Tree Seed Centre, Muak Lek, Saraburi, Thailand 18180 and † Centre for Forest Biology, Uniersity of Victoria, Victoria, B.C., Canada V8W 2Y2 Received : 28 August 1996
Accepted : 2 April 1997
Pollen released at 1100 h has the highest viability (92±2 %) but is no longer viable 3 d (84 h) after anthesis. In itro pollen-tube growth is fast (140 µm h−") and increases significantly within the first 8 h. In io pollen tubes also grow quickly and reach the base of the style within 2 h after pollination and enter the micropyle 8 h after pollination. There is no significant difference between self- and cross-pollination in either the rate and the number of pollen tubes in the pistil and the number of ovules penetrated by a pollen tube. Teak has late-acting gametophytic self-incompatibility ; the majority of pollen tubes grow through the style but some do not continue to grow from the style towards the embryo sacs. Pollen-tube abnormalities include swollen, reversed, forked and tapered tips and irregular and spiralling tubes. These are most prevalent in self-pollination (20±4 %). The index of self-incompatibility of 0±17 and low fruit set following self-pollination (2±49 %) indicates that teak is mostly self-incompatible. Drastic fruit abortion occurs within the first week following controlled pollination. Within 14 d, fruit size and fruit set from cross-pollination is generally much greater than from self-pollination. # 1997 Annals of Botany Company Key words : Tectona grandis, pollen viability, pollen-tube growth, pollination, controlled pollinations, incompatibility.
INTRODUCTION Like many hermaphroditic species (Stephenson, 1981 ; Willson and Burley, 1983 ; Sutherland, 1987), teak (Tectona grandis Linn. f.) produces low fruit set due to selfincompatibility (SI) (Hedegart, 1973 ; Tangmitcharoen and Owens, 1997). This is also true for the two other members of the Verbenaceae ; Vitex cooperi and Gmelina arborea (Bolstad and Bawa, 1982 ; Bawa, Perry and Beach, 1985). In teak, it was suggested that insect pollinators foraged mostly on the same tree resulting in a high level of selfing which led to lower fruit set and production of self-fruits with low germination rates (Hedegart, 1973). There have been several studies of the reproductive biology and pollination of teak (Cameron, 1968 ; Bryndum and Hedegart, 1969 ; Hedegart, 1973 ; Egenti, 1974 ; Kedarnath, 1974 ; Tangmitcharoen and Owens, 1997). Bryndum and Hedegart (1969) described the procedure for controlled pollination using small bags to isolate single flowers and large bags to isolate an entire inflorescence. Hand crosspollination was reported to increase the fertilization percentage by 6 to 60 % (average 20 %) compared to 0±4 to 5±1 % (average 1±3 %) from open-pollination. However, there has been no report on the pattern of pollen-tube growth or quantification of pollen tubes within the pistil following hand-pollinations and their role in fruit production of teak. To ensure the success of hand-pollinations, it is important to determine pollen viability prior to pollen application. In general, pollen viability indicates the ability to deliver the ‡ For correspondence. Fax 604 721 6611, e-mail : jowens!uvic.ca
0305-7364}97}10040110 $25.00}0
sperm cells to the embryo sac after pollination (Shivanna, Linskens and Cresti, 1991). In species having gametophytic self-incompatibility (GSI), pollen tends to be binucleate and germinate readily in itro (Brewbaker and Majumder, 1961 ; Brewbaker, 1967). Teak pollen is reported to be binucleate (Siripatanadilox, 1974) but speed or ease of germination are not known. Mulcahy and Mulcahy (1983) also suggested that pollen tubes from binucleate pollen tend to be inhibited in the style in GSI. In angiosperms, the rate of in itro pollen-tube growth varies from 60–20 000 µm h−". This is at least ten times faster than gymnosperm pollen (Hoekstra, 1983). Brewbaker and Majumder (1961) proposed that in itro growth of binucleate pollen tubes is approximately 10 % of in io pollen-tube growth due to limited reserve food in the pollen. However, Bryndum and Hedegart (1969) reported that teak in itro pollen-tube growth rate was high, reaching a maximum of 9±3 mm in 24 h. Self-incompatibility may occur in the stigma, style or ovary (Seavey and Bawa, 1986) but inhibition in the style is thought to be more characteristic of GSI (de Nettancourt, 1977). In open-pollination of teak (Tangmitcharoen and Owens, 1997) pollen-tube inhibition from GSI occurred anywhere along the pistil, but was mostly in the lower portion of the ovary. Incompatible pollen-tube arrest within the ovary, described as late-acting SI, has now been reported to occur primarily in woody species (Seavey and Bawa, 1986 ; Sage, Bertin and Williams, 1994). This study reports on pollen viability and longevity, the patterns and the rates of in itro and in io pollen-tube growth following controlled pollinations, various abnormalities observed in io pollen-tube growth, rate and form
bo960440
# 1997 Annals of Botany Company
402
Tangmitcharoen and Owens—Controlled Pollination and Fruit Production in Teak
of SI, reproductive success, fruit set and rate of fruit abortion of teak.
Three 40-year-old plantation teak trees growing at Muaklek, Saraburi, Thailand, (14° 40« N, 101° 17« E) were used. Two or three trees were used for each part of the study. Scaffolding was erected to a height of 8 m adjacent to each tree. Flowering occurs during the rainy season from June to August and the most abundant flower production occurs in July and August The study began in July 1994.
previously bagged flowers and was applied by tapping the anther, held with forceps, over the stigma of a receptive flower between 1100 and 1200 h. Pollen donors were chosen from trees located nearby (5 m) and further (500 m) away. For self-pollinations, pollen from the same tree was used and applied as described above. Since approx. 1 to 3 % of the flowers in a given inflorescence open each day (Tangmitcharoen and Owens, 1997), hand pollinations were carried out on successive days for each inflorescence. The bags were removed the day after all of the open flowers in the bag were pollinated to prevent inflorescence damage by heavy rains and strong winds against the bags.
Pollen germination and pollen-tube growth
Microscopy
Preliminary studies indicated that media containing only sucrose may not result in optimum pollen germination. Therefore, Brewbaker’s solution (Brewbaker and Kwack, 1963) was used for this study. To determine pollen germination, mature pollen was extracted at 12 different times (0, 2, 4, 6, 8, 10, 12, 24, 36, 48, 72, 84 h) after flower opening (0700 h). For each collection, two flowers per inflorescence (four inflorescences per tree) were collected from each of the three trees. The fresh pollen from each flower was then placed in a petri dish containing 10 % Brewbaker’s solution with 10 % sucrose for 4 h. A drop of this mixture was placed on a microscope slide and covered with a cover slip. Using a light microscope, approx. 200 pollen grains per sample were observed to determine the percentage and the time of pollen germination, i.e. when the pollen tubes were equal or greater in length than pollen diameter. To determine in itro pollentube growth, mature pollen collected at 1300 h was placed on a drop of the same solution and covered with a coverslip. Using an eyepiece gradicule, 90 germinated grains from the three trees were measured for pollen-tube length at 1, 2, 4, 8, 12 and 24 h after sowing.
Specimens used for transmission electron microscopy (TEM) were fixed in 2±5 % glutaraldehyde with 0±075 PO % buffer (pH 7±2) for 2 h at room temperature, rinsed in 0±075 PO buffer, then postfixed in 1 % osmium tetroxide % for 1 h. Specimens were dehydrated in an ethanol series, infiltrated with Spurr’s resin (Spurr, 1969) and cured for 18 h at 60 °C. Semithin sections (1 µm) for light microscopy (LM) were stained with Richardson’s stain (Richardson, Jarrett and Finke, 1960). Ultrathin sections were placed on uncoated 200-mesh copper grids, stained with 2 % aqueous uranyl acetate and 0±2 % lead citrate (Reynolds, 1963) and viewed with a JOEL JEM 1200 EX electron microscope.
MATERIALS AND METHODS Plant material and study site
Pollination experiments In a previous study (Tangmitcharoen and Owens, 1997), pollination success was greater in flowers which opened early and during the peak flowering season than in those which opened at the end of the flowering season. Therefore, hand pollinations were generally performed on those flowers which opened at the beginning or middle of the inflorescence flowering period. At pollination, each flower was tagged with coloured thread and one of the following pollination treatments was carried out : (1) flowers not emasculated, self-pollinated, and bagged ; (2) flowers emasculated, crosspollinated, and bagged ; and (3) flowers not emasculated and unbagged (open-pollination). In the emasculated treatments, the undehisced stamens were removed from the newly opened flowers between 0600 and 0700 h on the day of receptivity. Cellophane bags were then placed over the group of open flowers (approx. ten flowers per inflorescence) to exclude all potential pollinators. Each bag was removed from the receptive flowers for about 15 min for hand pollinations. The pollen was obtained from
Pollen-tube growth and self-incompatibility To determine in io pollen-tube growth, 15 flowers from each of five inflorescences from each of two trees were collected at 1, 2, 4, 6, 8, 12, 24 and 48 h following pollination. The pistils were sliced longitudinally to remove the hairy lateral surfaces of the ovary then fixed in ethanolacetic acid (3 : 1 v}v) for 24 h. After rinsing with water, they were cleared in 8 NaOH at room temperature for 2–4 d, or until most of the tissues became transparent. They were then rinsed in water and stained with 0±1 % decolorized aniline blue in 0±1 K PO modified from Dumas and Knox $ % (1983) and Kenrick and Knox (1985). The number and lengths of pollen tubes in the pistils were examined in approx. 350 flowers using epi-fluorescence microscopy. Fruit set and fruit size To determine fruit set and fruit size following controlled pollination, the variously pollinated flowers were observed and monitored every day for 1 month. The index of selfincompatibility (ISI) was determined by dividing the percentage of fruit set from self-pollination by the percentage of fruit set from cross-pollination (Zapata and Arroyo, 1978). Seed efficiency and reproductie success from openpollination Seed efficiency (SE) was calculated as the number of filled seeds, divided by seed potential, multiplied by 100. To
403
Tangmitcharoen and Owens—Controlled Pollination and Fruit Production in Teak
Statistical analysis Means³s.e. were calculated for all measurements. Arcsine transformations were applied to percentage of pollen germination, the percentage of pollinated flowers, percentage of fruit abortion, and distribution of number of seeds per fruit prior to analysis of variance (ANOVA). The among tree variation in in itro pollen germination, rate of in itro and in io pollen-tube growth, and rate of fruit abortion from controlled pollinations were assessed by ANOVA. The Duncan new multiple range test at P ! 0±05 was used to compare the means for a significant difference among the variables.
1400 d d
d
1200 Pollen tube length (micron)
determine filled seed per fruit, 450 fruits were collected and examined using the cutting tests and x-ray at 25 kv for 60 s. Reproductive success is the product of the fruit to flower ratio (Fr}Fl) and the seed to ovule ratio (S}O) (Weins et al., 1987). Ten inflorescences from each of the three trees were counted to obtain the total number of flowers. From the same inflorescences, the number of flowers that developed into mature fruits were counted.
1000
c
800 600 400
b
200 0
a
1
2
4 8 12 Time after sowing (h)
24
F. 2. In itro pollen-tube growth of teak up to 24 h after sowing. Vertical bars represent s.e. Means of each variable with the same letter are not significantly different at P ! 0±05 as determined by the Duncan new multiple range test.
RESULTS In vitro pollen germination and pollen-tube growth
60
50
40 % of flowers
Pollen released at 1100 h (4 h after anthesis) had the highest viability (92±2 %). Pollen viability gradually decreased after 1100 h and 3 d (84 h) after flower opening, pollen was no longer viable (Fig. 1). There was no significant difference in pollen germination among trees. The peak receptive period for teak flowers was 1100 to 1300 h as reported in an earlier study (Tangmitcharoen and Owens, 1997). Pollen germination began within 1 h of being placed in
30
20
100
a b
b
10
d
34–36
31–33
28–30
25–27
22–24
19–21
16–18
d
d
13–15
60
10–12
0
7–9
c
4–6
c
1–3
% pollen germination
80
Pollen grains per stigma F. 3. Percentage of flowers having different numbers of pollen grains on the stigma of control-pollinated teak flowers ; self (*), n ¯ 49 ; open (+), n ¯ 54 ; cross (8), n ¯ 29.
40 c 20 f f 0
0
2
g
4 6 8 10 12 24 36 48 72 84 Time after anther opening (h)
F. 1. In itro pollen germination of teak pollen at various times after anther opening (n ¯ 144). Vertical bars represent s.e. Means of each variable with the same letter are not significantly different at P ! 0±05 as determined by the Duncan new multiple range test.
Brewbaker’s solution. Pollen-tube growth was fast and increased significantly each hour from 0 to 8 h. Within the first 2 h, the average rate of tube growth was about 160 µm h−". During the 3 to 4 h period it was 280 µm h−". Overall, the rate of tube growth within the first 8 h varied from 60 to 280 µm h−" with an average of 140 µm h−" (Fig. 2). After 8 h, there was no significant increase in pollen-tube growth. The rate of pollen-tube growth varied among trees (P ! 0±001).
404
Tangmitcharoen and Owens—Controlled Pollination and Fruit Production in Teak The average number of pollen tubes in the style varied from 5 to 8. The majority of pollen tubes grew through the style (Fig. 6) but some did not continue to grow from the style towards the embryo sacs (Fig. 7). Pollen-tube growth was fast during the first 4 h. Four h after pollination, pollen tubes of 24 % of self-, 29 % of openand 8 % of cross-pollinated flowers had reached an ovule and about 7 % of pollen tubes from open-pollinated plants had reached a micropyle (Fig. 5 A). The first pollen tubes were observed to reach the embryo sacs 8 h after pollination (Fig. 5 B). Pollen-tube growth was slow between the 8 to 12 h and the 20 to 24 h observation periods. At the 20 to 24 h observation period, 70 to 80 % of pollen tubes from the three types of pollination had reached the ovules, 40 to 60 % of pollen tubes had reached the micropyle, but very few (! 5 %) of pollen tubes from self- and open-pollinated flowers and about 20 % from cross-pollinated plants had reached the embryo sacs (Fig. 5 C). The number of pollen tubes at different positions within the pistil at 20 to 24 h following controlled pollination is shown in Table 1. Of the 84 flowers observed 8 to 24 h after pollination, the number of flowers in which pollen tubes entered the micropyle did not differ significantly among the three types of pollination. The average number of pollen tubes in the various portions of the pistil 20 to 24 h after pollination is shown in Table 1. The number of pollen tubes entering the micropyle per flower varied from 0 to 3 but was not significantly different among the three types of pollination (Table 1). In general, only one micropyle per flower was entered by a pollen tube even though four micropyles exist per flower. Occasionally two, very rarely three and never four micropyles per flower were observed to contain pollen tubes. Observation of the pistil at 28, 32, 36 and 48 h after pollination indicated that there was no increase in the number of pollen tubes entering the micropyles after 24 h.
Stigmatic surface
Pollen Pollen tubes
Style
Ovary Ovule Embryo sac
F. 4. Diagram of teak flower showing the pathway of pollen tubes through the stigma and hollow stylar transmitting tissue (shaded area) to the embryo sacs.
In vivo pollen-tube growth following controlled pollination Of the 225 flowers observed following control self-, openand cross-pollinations, 30 to 50 % (average 35 %³4±16) had pollen adhering to the stigma. The number of pollen grains on the stigma varied from 1 to 38 with an average of 6±2³0±63 (n ¯ 130), but most commonly there were only 1 to 3 grains (Fig. 3). The percentage of pollen adhering to the stigma did not differ significantly between the three types of pollination. The pathway of pollen-tube growth to the embryo sacs is shown in Fig. 4. In all three types of pollination, a few pollen grains did not germinate on the stigma (Table 2). Pollen germinated within 1 h of being placed on a receptive stigma. Generally, self-, open-(natural), and cross-pollen tubes grew at the same rate and there were similar numbers of pollen tubes growing through the pistil for the first 24 h.
Abnormalities of pollen-tube growth In the 132 pistils observed during the 24 h following controlled pollination, most pollen tubes were straight with thin pollen-tube walls and tapered tips. Pollen-tube inhibition on the stigma occasionally occurred in all types of pollination (Fig. 8). Several abnormalities occurred in pollen-tube growth at various positions within the pistil (Table 2). Abnormalities included reversing tubes (Fig. 9),
T 1. Comparison of mean numbers (³s.e.) of pollen tubes entering the stigma, at 25, 50, 75 and 100 % of the style length, and in the oule, micropyle and embryo sac at 20–24 h following pollinations Types of pollination
ANOVA
Distance through pistil
Self-
Open-
Cross-
F values
P
Average
25 % of style 50 % of style 75 % of style 100 % of style Ovule Micropyle Embryo sac
7±8³2±0 7±7³2±0 7±3³2±0 7±3³2±0 6±6³1±9 3±9³1±3 0±2³0±09
6±6³2±5 6±4³2±5 6±4³2±5 6±2³2±5 6±0³2±4 4±3³2±0 0±5³0±22
5±2³1±1 4±9³1±1 4±9³1±1 4±9³1±1 4±6³1±2 2±0³0±63 0±5³0±14
0±45 0±52 0±50 0±39 0±28 0±70 1±28
0±64 NS 0±60 NS 0±62 NS 0±68 NS 0±76 NS 0±50 NS 0±29 NS
6±7 6±6 6±4 6±3 5±8 3±5 0±3
NS, non significant.
Tangmitcharoen and Owens—Controlled Pollination and Fruit Production in Teak 100 2–4 h
A
80 60 40 20 0 25%
50%
75%
100%
ov
mc
em
405
by the percentage fruit set from cross-pollination (14±54) was 0±17, indicating that teak is mostly self-incompatible (Zapata and Arroyo, 1978). Fruit set for self- and openpollinated flowers measured 30 d after pollination differed significantly (P ! 0±01) from cross-pollinated flowers. Only 2±49 % of self-pollinated flowers and 6±54 % of openpollinated flowers set fruit, whereas 14±54 % of crosspollinated flowers set fruit (Table 3). There was no significant difference in percent fruit set among the two trees used for this portion of the study for all types of pollination. The rate of fruit abortion was high during the first 10 d. There was a tendency for a lower percentage of fruit abortion following cross-pollination during the 5 to 20 d period after pollination (Fig. 13).
100 Percent of pollen tubes
8–12 h
B
80
Seed efficiency and reproductie success with openpollination
60
In open-pollination, the seed to ovule (S}O) ratio was 0±33 and the fruit to flower (Fr}Fl) ratio was 0±035. Reproductive success, determined by Fr}Fl multiplied by S}O, was low (0±011). The results of cutting tests and x-ray revealed that most commonly (46±2 %) only one of four ovules per flower developed into a mature seed, 30±9 % contained two seeds per fruit, 8 % contained 3 to 4 seeds and 14±9 % of fruits contained no filled seeds (Table 4). There was no significant difference in seeds per fruit among the three trees.
40 20 0 25%
50%
75%
100%
ov
mc
em
100 20–24 h
C
80
Fruit size during early deelopment stages Fruit size during early development, up to 10 d after pollination, differed significantly (P ! 0±0001) among the types of pollination but not among the two trees included in this part of the study. Self-pollination resulted in smaller fruits. Unfortunately, the high abortion rate due to selfing made fruit size data unavailable after 10 d. Mean fruit size resulting from open- and cross-pollination was essentially the same through the first 28 d of development (Table 5).
60 40 20 0 25%
50%
75% 100% ov mc Distance through style
em
F. 5. Percentage of self- (*), open- (natural) (+), and cross- (8) pollen tubes at different positions in the teak pistil after pollination. A, At 2–4 h, self- (n ¯ 14) ; open- (n ¯ 14) ; cross- (n ¯ 4). B, At 8–12 h, self- (n ¯ 14) ; open- (n ¯ 31) ; cross- (n ¯ 11). C, At 20–24 h, self(n ¯ 17) ; open- (n ¯ 45) ; cross- (n ¯ 9). 25, 50, 75 and 100 % refer to percent of the style length ; ov, ovule ; mc, micropyle ; em, embryo sac.
irregular or spiralling tubes (Fig. 10), and an increase of callose deposits resulting in swelling of the tube tip (Figs 11 and 12). Abnormalities were most prevalent in the selfpollinated flowers (20±4 %) (Table 2). Swelling of the tube tip was most common, accounting for 41±7 % of total abnormal tubes. Fruit set and rate of fruit abortion following controlled pollination The index of self incompatibility (ISI), determined by dividing the percentage fruit set from self-pollination (2±49)
DISCUSSION In vitro pollen germination Binucleate pollen generally germinates readily in culture (Mulcahy and Mulcahy, 1983). This was also true for binucleate teak pollen in which germination began within 1 h of pollen being placed in the medium. Using a 14 % sucrose medium for in itro pollen germination of teak, Egenti (1974) and Bryndum and Hedegart (1969) reported that pollen collected during the 1200–1400 h period on the day of anthesis had the highest percentage of viability compared to that collected from 0730–1000 h and at 1800 h. They also found that 10 % of the pollen was still viable 3 d after anthesis. Using Brewbaker’s solution with 10 % sucrose, results from our studies show some differences to those of previous studies. Germination was best for pollen collected 1 h earlier (1100 h) than in previous studies, but the pollen also remained viable for 3 d.
406
Tangmitcharoen and Owens—Controlled Pollination and Fruit Production in Teak
F 6 and 7. Epi-fluorescence micrographs of pollinated teak pistils within 24 h of pollination. Pistils were stained with decolorized aniline blue to localize callose. Fig. 6. Within 4 h, a large number of pollen germinate on the stigma papillae and produce pollen tubes that grow into the stylar canal (arrowheads). Fig. 7. Arrested pollen tubes in the ovary 24 h after pollination. A dense bundle of pollen tubes (between arrowheads) was generally seen in the lower stylar transmitting tissue but few grew towards the lower portion of the ovary and entered the micropyles. Bars ¯ 0±5 and 0±2 mm in Figs 6 and 7, respectively. P, Pollen ; OV, ovule. F 8–12. Abnormalities of pollen tubes found at various sites and times following pollination. Fig. 8. Inhibition of a pollen tube on the stigma 12 h after self-pollination. Fig. 9. Reversal of pollen-tube growth (arrowheads) in the style 4 h after self-pollination. Fig. 10. Irregular pollen tube with callose deposition in pollen tube wall (arrowhead) in the upper portion of the ovary 8 h after open-pollination. Fig. 11. Light micrograph showing swelling of pollen tube (arrowhead) in the lower portion of the ovary about 24 h after self-pollination. Fig. 12. Transmission electron micrograph showing the thickened cell wall and small amount of cytoplasm (*) of the tube tip shown in Fig. 11. Bars ¯ 50 mm in Fig. 8 ; 0±3 mm in Figs 9 and 10 ; 20 µm in Fig. 11 ; and 10 µm in Fig. 12. P, Pollen ; ST, stigma ; PT, pollen tube ; OV, ovule ; OW, ovary wall.
In general, most pollen collected on the day of anthesis developed normal pollen tubes and exhibited more uniform pollen-tube growth than pollen collected on or after the
second day. The highest in itro pollen germination (92±2 %) occurred in pollen released at 1100 h. This indicates that the pollen used for control pollination was highly viable.
407
Tangmitcharoen and Owens—Controlled Pollination and Fruit Production in Teak T 2. Number of flowers that had different abnormalities of pollen-tube growth on the stigma and at arious positions within the pistil within 24 h following controlled pollination
100
80
Types of pollination
Stigma* 50 % of style 75 % of style 100 % of style Upper ovary Lower ovary Total number of flowers having abnormal pollen tubes Total number of flowers observed
Self-
Open-
Cross-
4 3 2 0 1R 1R 2S 1R 0 1R 0 0 1R 0 0 1I 3S 1I 0 10 6 5 (20±41 %) (11±11 %) (17±24 %) 49 54 29
* Stigma with pollen adhering but without pollen tube growth. R, Reversing tubes ; S, swollen tube tip ; I, irregular tube.
T 3. Number and mean (³s.e.) percentage of fruit set 30 d after controlled pollination
% Fruit abortion
Distance through pistil
60
40
20
0
5
10 15 20 Days after pollination
25
30
F. 13. Effect of self- (+), open- (D), an cross- (^) pollination on fruit abortion for the 30 d after pollination based on 14 inflorescences on each of the two teak trees.
Types of pollination Tree number Number of inflorescences treated Number of flowers treated Number of fruits set % fruit set Total (% fruit set)
1 2 1 2 1 2 1 2
Self-
Open-
Cross-
3 4 6 4 5 6 30 126 67 52 109 75 1 2 10 1 10 12 3±33 1±59 14±82 1±92 9±17 15±54 2±49a 6±54a 14±54b ³1±61 ³4±65 ³2±77
Means of each variable followed by the same superscript are not significantly different (P ! 0±05) as determined by the Duncan new multiple range test.
Therefore, most of the failure of pollen-tube growth which occurred after control pollination was not due to the lack of pollen vigour and viability, but probably due to incompatibility. In vitro and in vivo pollen-tube growth in controlled pollination There are extreme variations in in io and in itro pollentube growth rate in angiosperms (Hoekstra, 1983). However, there are few reports of in io or in itro pollen-tube growth in tropical forest trees. Rapid in io pollen-tube growth has been reported in Acacia retinotes in which the pollen tubes reached the ovule within 11 h (Kenrick and Knox, 1985). In other cases growth is slow ; it took 10–20 d for pollen tubes of Eucalyptus woodwardii (Sedgley and Smith, 1989) to reach the ovules. In our study, the average rate of in itro pollen-tube growth varied from 60–280 µm h−" which is similar to the majority (19 of the 26) of angiosperm families
T 4. Frequency distribution of percentage fruit haing 0–4 seeds per fruit 30 d after pollination Numbers of seeds per fruit 0
1
2
14±9a 46±2b 30±9c 67 208 139 3±5 3±7 4±3
Percentage of fruits Numbers of fruits s.e.
3 6±22d 28 2±1
4
Total
1±78d 100 8 450 0±69
Means of each variable followed by the same superscript are not significantly different (P ! 0±05) as determined by Duncan new multiple range test.
T 5. Comparison of means (³s.e.) of fruit sizes in mm during early deelopment Types of pollination Days after pollinations 4 7 10 14 21 28
Self-
Open-
Cross-
2±25³0±07 (20)a 2±36³0±06 (20)b 2±45³0±08 (19)b 2±42³0±09 (20)a 3±86³0±12 (20)b 4±2³0±58 (17)b 3±07³0±09 (14)a 5±71³0±13 (19)b 6±41³0±16 (14)b N}A 7±03³0±21 (18) 7±46³0±73 (16) N}A 12±09³0±49 (15) 12±94³0±58 (14) N}A 14±15³0±33 (11) 14±17³0±34 (13)
Numbers in brackets represent the number of fruits observed. Means of each variable followed by the same superscript are not significantly different (P ! 0±05) as determined by the Duncan new multiple range test. N}A, Data not available due to high abortion rate followed selfpollination.
reported by Hoekstra (1983). In our study pollen tubes reached about 1±2 mm in 24 h. This contrasts with studies by Bryndum and Hedegart (1969) for teak where in itro
408
Tangmitcharoen and Owens—Controlled Pollination and Fruit Production in Teak
pollen-tube growth was very fast, up to 9±3 mm with an average of 5±53 mm in 24 h. Although the rate of in itro pollen-tube growth in teak would be considered fast, it was not when compared to the length of pistil. Pistil length averaged 6±55 mm and consequently 24 h after sowing pollen tubes in itro were only about 18 % of the way down the pistil. The in itro growth rates underestimate the in io rates in teak in which some pollen tubes reached the ovules 4 h after pollination. Our results agree with the study by Brewbaker and Majumder (1961) that in itro growth of binucleate pollen tubes was approx. 10 % of in io pollentube growth. Rosen and Gawlick (1966) and Rosen (1971) proposed that in io binucleate pollen-tube growth consists of two phases : autotrophic and heterotrophic. In the first autotrophic phase pollen tubes grow from their own reserves. Pollen grains usually contain enough reserve food to support germination (Johri, 1992). Pollen-tube growth during the first phase is relatively slow and free of callose plugs (Mulcahy and Mulcahy, 1983). Brewbaker (1967) suggested that this phase ends with gamete formation. Pollen-tube growth in itro, or in an incompatible style, may terminate at this point. The second, heterotrophic, phase is observed only in io within a compatible style. During this phase, pollen tubes grow rapidly and form callose plugs, indicating a shift to heterotrophic nutrition (uptake of substances from the pistil). This may account for the limited teak pollentube length in itro using Brewbaker’s media. It may not contain all of the nutrients needed for the transition of pollen tubes to the second phase (Read, Bacic and Clarke, 1992). The continual rapid pollen-tube growth in the stylar canal of teak may be related to movement of the specific proteins from the transmitting tissue, as suggested by Mascarenhas (1993). Hohri (1992) also suggested that a pollen tube growing in a hollow style, as in teak, derives nourishment from the glandular epidermis of the stylar canal and the adjacent tissue becomes depleted of metabolites. Despite these complications, in itro pollen-tube growth measurements reflect some useful pollen properties (Hoekstra, 1983). Based on our observations of in io pollen-tube growth, it is likely that the period for the first phase of in io pollentube growth in teak was very short (probably less than 30 min). Within 1 h following pollination, callose plugs were found near the stigma which corresponds to first phase of growth. The density of pollen deposited on the stigma is reported to affect both pollen germination and rate of pollen-tube growth in some tropical forest trees, including Leucaena leucocephala (Ganeshaiah, Shaankar and Shivashankar, 1986). Our study showed that this may not be true for teak. The density of pollen deposited on the teak stigma did not affect the number of pollen tubes entering the micropyle in any pollination trials. Even when the teak stigma was saturated with pollen (up to 38 grains per stigma), the ovule often remained unpenetrated. There was little pollen competition in the teak style, i.e. an increase in pollen load on the stigma resulted in a proportional increase in the number of pollen tubes penetrating to the base of the ovary.
Pollen-tube inhibition and incompatibility in teak There was no difference in the number or sites of pollentube inhibition among open-, self-, and cross-pollinations in teak until pollen tubes reached the embryo sac (Fig. 5). Also, in the previous report on open-(natural) pollination for teak (Tangmitcharoen and Owens, 1997), the site of the final pollen-tube arrest was generally at the ovary, and most pollen tubes did not enter the embryo sac. Other studies in Eucalyptus woodwardii (Sedgley and Smith, 1989) and Medicago satia (Cooper and Brink, 1940 ; Sayers and Murphy, 1966) have shown that the number of penetrated ovules following self-pollination was less than that following cross-pollination. In angiosperms, self-incompatibility (SI) systems that operate in the ovary, described as late-acting SI systems, were initially assumed to be uncommon in angiosperms (de Nettancourt, 1977). Seavey and Bawa (1986) reported that this notion was erroneous : the scarcity of late-acting SI may have been exaggerated by the preference for studies of breeding systems for small, short-lived, herbaceous plants. Late-acting SI has been reported to be important in the breeding systems of many angiosperms and may be more important in woody angiosperms (Seavey and Bawa, 1986 ; Sage et al., 1994). They were classified into four categories : (1) ovarian inhibition of incompatible pollen tubes before the ovule is reached ; (2) pre-fertilization inhibition in the ovule ; (3) post-zygotic rejection of the embryo ; and (4) ovular inhibition. Late-acting pollen-tube inhibition has been reported in some hardwood species ; Acacia retinodes (Kenrick, Kaul and Williams, 1986), Eucalyptus morrisbyi (Potts and Savva, 1989), and E. woodwardii (Sedgley and Smith, 1989). In these, self- and cross-pollen tubes appear similar through the stigma and style, but self-pollen tubes were inhibited in the ovary. In teak, the late-acting system appears to be associated with inhibition after pollen tubes enter the ovule, however, a post-zygotic stage has not been ruled out. Post-zygotic outcrossing mechanisms have been reported in some tropical woody species including four species of Eucalyptus (Griffin, Moran and Fripp, 1987 ; Ellis and Sedgley, 1992) and mango (Shama and Singh, 1970). In all pollination trials, a variety of pollen-tube abnormalities were found at various locations in the pistil and at different times after pollination. Swollen tips, reversing tips, forked tips, tapered tips, irregular tubes and spiralling tubes were observed. There was a tendency towards higher pollen-tube abnormalities in self-pollinated flowers than open- and cross-pollinated flowers (Table 2). Fruit set following controlled pollinations Low fruit set is typical of many hermaphroditic plant species (Stephenson, 1981 ; Willson and Burley, 1983 ; Sutherland and Delph, 1984 ; Sutherland, 1986, 1987 ; Guitian, 1993). Zapata and Arroyo (1978) classified the ISI as : " 1 ¯ self-compatible ; " 0±2 ! 1 ¯ partially selfincompatible ; ! 0±2 ¯ mostly self-incompatible ; and 0 ¯ completely self-incompatible. The index of ISI was 0±17 in teak, placing it in the category of mostly self-incompatible. This helps explain the high rate of fruit abortion that occurs
Tangmitcharoen and Owens—Controlled Pollination and Fruit Production in Teak after self-pollination, but some self-pollinated flowers can still develop into fruits. Dafni (1992) ranked fruit set percent following selfpollination as : 0–3 % (class 0) ¯ self-incompatible ; 3–30 % (class 1) ¯ slightly self-compatible ; and, " 30 % (class 2) ¯ highly self-compatible. Our study shows the rate of fruit set following self-pollination in teak was 2±49 %, which again classifies teak as a self-incompatible species. Fruit set did not differ significantly between self- and open-pollination which suggests a high incidence of self-pollination in nature or very closely related neighbouring trees. Results from our study support an earlier study by Bryndum and Hedegart (1969) in which less than 1 % of self-pollinated flowers developed into fruits. Evidence from the present and our earlier study (Tangmitcharoen and Owens, 1997) show that teak has late-acting GSI, where the pollen-pistil interaction is genetically controlled by the haploid (gametophytic) genome of each pollen tube as it penetrates the diploid pistil (Sedgley and Griffin, 1989). Low fruit set has also been reported in two other species from the Verbenaceae, Vitex cooperi and Gmilina aborlia (Bolstad and Bawa, 1982 ; Bawa, Perry and Beach, 1985), but in these it was not necessarily attributed to self-incompatibility. In teak, drastic fruit abortion occurred within the first week following control pollination and within 14 d fruit size and fruit set from cross-pollination was generally much greater than from self-pollination. Ackerman (1961) reported that fruit set from self-pollination of Ziziphus (chinese jujubes) was less and fruit had a greater tendency to drop prematurely than from cross-pollination. In general, an increase in pollen load by supplementary pollination leads to an increase in fruit set (Sutherland and Delph, 1984 ; Sutherland, 1987). This was shown in macadamia where an increased number of pollen tubes following crosspollination increased initial fruit set (Ito, Eyre and Cabral, 1983 ; Sedgley et al., 1990). Insufficient insect pollinators and their effectiveness appear to be major causes for limited fruit set in teak (Hedegart, 1973 ; Tangmitcharoen and Owens, 1997). Insects forage mostly among the inflorescences of the same tree or nearby relatives, which contribute to inbred fruits (Hedegart, 1973). This results in a lack of heterozygosity, low fruit set and low germination rate in teak. However, Kertadikara and Prat (1995) suggested that there is high heterozygosity in teak populations and it may be maintained by early elimination of self-material through embryo abortion, and low germination rate. Low fruit set after cross-pollination in many tropical forest species probably results from three factors (Bawa et al., 1985) : an artifact of hand-pollination (Bawa et al., 1985) ; a predetermined abortion rate (Bawa and Webb, 1984) ; and inbreeding depression from close relative crosspollination (Haber and Frankie, 1982). In our study these factors may have contributed to the relatively high (85±46 %) fruit abortion of teak with cross-pollination. Even in crosspollinated flowers, only 30³6±1 % (n ¯ 95) of the flowers had pollen on the stigma and fruit abortion occurred more often if hand-pollination was performed on the flowers which developed at the end of the flowering period in an inflorescence, in particular the outermost flowers of the
409
inflorescence. Also the trees used in our study may be close relatives since they were located in the same plantation. A C K N O W L E D G E M E N TS This study was a part of M.Sc. research of Tangmitcharoen Suwan, funded by CIDA (Canadian Development Agency) with the Canadian Forest Service (Petawawa National Forestry Institute) as Canadian Executing Agency to ASEAN Canada Tree Seed Centre, Muak Lek, Saraburi, Thailand. Research support was also provided by the National Sciences and Engineering Council of Canada, Grant 1982 to JNO. We thank the ASEAN Forest Tree Seed Centre, Saraburi, Thailand for providing field and laboratory facilities. We also thank the Thai Danish Dairy Farm, Saraburi, Thailand for providing the collection site. LITERATURE CITED Ackerman WI. 1961. Flowering, pollination, self-sterility and seed development of chinese jujubes. Proceedings of the American Society for Horticultural Science 77 : 265–269. Bawa KS, Perry DR, Beach JH. 1985. Reproductive biology of tropical lowland rain forest trees. I. Sexual systems and incompatibility mechanisms. American Journal of Botany 72 : 331–345. Bawa KS, Webb CJ. 1984. Flower, fruit and seed abortion in tropical forest trees : implications for the evolution of paternal and maternal reproductive patterns. American Journal of Botany 71 : 736–751. Bolstad PV, Bawa KS. 1982. Self-incompatibility in Gmelina arborea L. (Verbenaceae). Silae Genetica 31 : 19–21. Brewbaker JL. 1967. The distribution and phylogenetic significance of binucleate and trinucleate pollen grains in the angiosperms. American Journal of Botany 54 : 1069–1083. Brewbaker JL, Kwack BH. 1963. The essential role of calcium ion in pollen germination and pollen tube growth. American Journal of Botany 50 : 859–865. Brewbaker JL, Majumder SK. 1961. Incompatibility and the pollen grain. Recent Adances in Botany 2 : 1503–1508. Bryndum K, Hedegart T. 1969. Pollination of teak (Tectona grandis L.f.). Silae Genetica 18 : 77–80. Cameron AL. 1968. Forest tree improvement in New Guinea, 1. Teak. 9th Commonwealth Forest Conference. New Delhi. Cooper DC, Brink RA. 1940. Partial self-incompatibility and collapse of fertile ovule as factors affecting seed formation in alfalfa. Journal of Agricultural Research 60 : 453–472. Dafni A. 1992. Pollination ecology. A practical approach. New York : Oxford University Press. de Nettancourt D. 1977. Incompatibility in angiosperms. Berlin, Heidelberg, New York : Springer-Verlag. Dumas C, Knox RB. 1983. Callose and determination of pistil viability and incompatibility. Theoretical and Applied Genetics 67 : 1–10. Egenti LC. 1974. Preliminary studies on pollinators of teak (Tectona grandis L.f.). Research Paper Forest Series, Federal Department of Forest Research, Nigeria. Ellis MF, Sedgley M. 1992. Floral morphology and breeding system of three species of Eucalyptus, Section Bisectaria (Myrtaceae). Australian Journal of Botany 40 : 249–262. Ganeshaiah KN, Shaankar RU, Shivashankar G. 1986. Stigmatic inhibition of pollen grain germination—its implication for frequency distribution of seed number in pods of Leucaena leucocephala (Lam) de Wit. Oecologia 70 : 568–572. Griffin AR, Moran GF, Fripp YJ. 1987. Preferential outcrossing in Eucalyptus regnans F. Muell. Australian Journal of Botany 35 : 465–475. Guitian J. 1993. Why Prunus Mahaleb (Rosaceae) produces more flowers than fruits. American Journal of Botany 80 : 1305–1309. Haber WA, Frankie GW. 1982. Pollination of Leuhea (Tiliaceae) in Costa Rican deciduous forest : significance of adapted and nonadapted pollinators. Ecology 63 : 1740–1750.
410
Tangmitcharoen and Owens—Controlled Pollination and Fruit Production in Teak
Hedegart T 1973. Pollination of teak (Tectona grandis L.). Silae Genetica 22 : 124–128. Hoekstra FA. 1983. Physiological evolution in angiosperm pollen : possible role of pollen vigour. In : Mulcahy DL, Ottaviano, eds. Pollen : Biology and implications for plant breeding. New York : Elsevier Science. Ito PJ, Eyre JM, Cabral S. 1983. Preliminary study on pollination of five macadamia cultivars. In : Stephenson RA, Gallagher EC, eds. Proceedings of the First Australian Macadamia Research Workshop. Brisbane : Australian Macadamia Society Limited. Johri BM. 1992. Haustorial role of pollen tubes. Annals of Botany 70 : 471–475. Kedarnath S. 1974. Genetic improvement of forest tree species in India. India Journal of Genetics Proceedings of 2nd General Congress Sarrao, New Delhi 34A. Kenrick J, Kaul V, Williams EG. 1986. Self-incompatibility in Acacia retinodes : Site of pollen-tube arrest in the nucellus. Planta 169 : 245–250. Kenrick J, Knox RB. 1985. Self-incompatibility in nitrogen-fixing tree, Acacia retinodes : quantitative cytology of pollen tube growth. Theoretical and Applied Genetics 69 : 481–488. Kertadikara AWS, Prat D. 1995. Genetic structure and mating system in teak (Tectona grandis L.f.) provenances. Silae Genetica 44 : 2–3. Mascarenhas JB. 1993. Molecular mechanisms of pollen tube growth and differentiation. Plant Cell 5 : 1303–1314. Mulcahy GB, Mulcahy DL. 1983. A comparison of pollen tube growth in bi- and trinucleate pollen. In : Mulcahy DL, Ottaviano E, eds. Pollen : Biology and implications for plant breeding. New York : Elsevier Science. Potts BM, Savva M. 1989. Self-incompatibility in Eucalyptus. In : Knox RB, Singh MB, Troiani, eds. Pollination ’88. University of Melbourne, Parkville : Biology Research Centre, pp. 165–170. Read M, Bacic A, Clarke AE. 1992. Pollen tube growth in culture. I. Control of morphology and generative nucleus division in cultured pollen tubes of Nicotiana. In : Ottaviano E et al., eds. Angiosperm pollen and oules. New York : Springer-Verlag, 162–167. Reynolds ES. 1963. The use of lead citrate at high pH as an electron opaque stain in electron microscopy. Journal of Cell Biology 17 : 208–212. Richardson KC, Jarrett L, Finke EH. 1960. Embedding in epoxy resins for ultrathin sectioning in electron microscopy. Stain Technology 35 : 313–323. Rosen WG. 1971. Pistil-pollen interactions in Lilium. In : HeslopHarrison J, ed. Pollen deelopment and physiology. London : Butterworths, 239–254. Rosen WG, Gawlick SR. 1966. Relation of lily pollen tube fine structure to pistil compatibility and mode of nutrition. Sixth International Congress for Electron Microscopy, Kyoto. Maruzen, Tokyo. Electron Microscopy 2 : 313–314.
Sage TL, Bertin RI, Williams EG. 1994. Ovarian and other late-acting self-incompatibility system. In : Williams et al., eds. Genetic control of self-incompatibility and reproductie deelopment in flowering plants. The Netherlands : Kluwer Academic Publishers, 116–140. Sayers ER, Murphy RP. 1966. Seed set in alfalfa as related to pollen tube growth, fertilization frequency, and post-fertilization ovule abortion. Crop Science 6 : 365–368. Seavey SR, Bawa KS. 1986. Late acting self-incompatibility in angiosperms. Botanical Reiew 52 : 195–219. Sedgley M, Bell FDH, Bell D, Winks CW, Pattison SJ, Hancock TW. 1990. Self- and cross-compatibility of macadamia cultivars. Journal of Horticultural Science 65 : 205–213. Sedgley M, Griffin AR. 1989. Sexual reproduction of tree crops. London : Academic Press, 378. Sedgley M, Smith RM. 1989. Pistil receptivity and pollen tube growth in relation to the breeding system of Eucalyptus woodwardii (Symphyomyrtus : Myrtaceae). Annals of Botany 64 : 21–31. Sharma DK, Singh RN. 1970. Self-incompatibility in mango (Mangifera indica L.). Horticultural Research 10 : 108–118. Shivanna KR, Linskens HF, Cresti M. 1991. Pollen viability and pollen vigor. Theoretical and Applied Genetics 81 : 38–42. Siripatanadilox S. 1974. Developmenmt of teak flower (Tectona grandis Linn.). Forest Research Bulletin 31 : Faculty of Forestry, Kasetsart University, Thailand. Spurr AR. 1969. A low viscosity epoxy resin embedding medium for electron microscopy. Journal of Ultrastructure Research 71 : 173–184. Stephenson AG. 1981. Flower and fruit abortion : proximate causes and ultimate functions. Annual Reiew of Ecology and Systematics 12 : 253–280. Sutherland S. 1986. Florax sex ratios, fruit set, and resource allocation in plants. Ecology 67 : 991–1001. Sutherland S. 1987. Why hermaphroditic plants produce many more flowers than fruits : experimental tests with Agae mckeleyana. Eolution 41 : 750–759. Sutherland S, Delph LF. 1984. On the importance of male fitness in plants : patterns of fruit-set. Ecology 65 : 1093–1104. Tangmitcharoen S, Owens JN. 1997. Floral biology, pollination, pistil receptivity, and pollen tube growth of teak (Tectona grandis L.f.). Annals of Botany 79 : 227–241. Wiens D, Calvin CL, Wilson CA, Davern CI, Frank D, Seavey SR. 1987. Reproductive success, spontaneous embryo abortion, and genetic load in flowering plants. Oecologia 71 : 501–509. Willson M, Burley N. 1983. Mate choice in plants : Tactics, mechanisms, and consequences. Princeton, New Jersey : Princeton University Press. Zapata TR, Arroyo MTK. 1978. Plant reproductive ecology of a secondary deciduous tropical forest in Venezuela. Biotropica 40 : 221–230.